The Agent Skills Directory

[COMMAND_EXECUTION]: The audit script executes system commands to verify the status of security-critical services.
Evidence: Uses subprocess.run to execute systemctl is-active for auditd, rsyslog, and syslog-ng in scripts/agent.py to confirm logging compliance.
[DATA_EXFILTRATION]: The skill accesses sensitive system configuration files to perform required compliance audits.
Evidence: Reads /etc/passwd to identify default or insecure accounts in scripts/agent.py (PCI Requirement 2.2.2).
Evidence: Reads PAM and password quality configuration files (/etc/pam.d/common-password, /etc/pam.d/system-auth, /etc/security/pwquality.conf) in scripts/agent.py to audit password complexity (PCI Requirement 8.3.6).
[INDIRECT_PROMPT_INJECTION]: The skill ingests data from local system files and network port scans, representing a potential injection surface if the environment is already compromised.
Ingestion points: Reads local system user data and security configuration files in scripts/agent.py.
Boundary markers: None defined; the script parses system files directly to extract configuration values.
Capability inventory: Performs network socket connections and executes system commands via subprocess.run.
Sanitization: Uses regular expressions to extract specific values from audited configuration files.

implementing-pci-dss-compliance-controls